WD-AccessData-1.0-20140312

[Douglas Tody]

Instruments that observe in multiple spectral bands simultaneously
produce cubes (or possibly multiple subarray-type Images) that identify
the image planes or subarrays by name rather than numerically.  The
bandpass names could be filter names, spectral line names, etc.  It is
similar to the polarisation case where the image planes are referred to
by name rather than numerically.  This is important mainly for
extracting named bandpass elements (subimages or subcubes), but could
also be relevant for discovery.  In the case of data access (either
accessData or automated virtual data generation), one might want to
retrieve the entire dataset, or an individual subimage or subcube.

To some extent the accessData problem can be avoided by requiring that
the spectral bands be exposed as separate virtual data products, however
the named/numerical problem will still be there, plus we need to use
association to relate the data products, metadata is duplicated, etc.

This use case is important because while it is theoretically possible to
do all this numerically, it can be hard, and is certainly inconvenient.
(This is why we had both named and numerical bandpass elements in SSA,
despite the additional complexity.)

 	- Doug


On Thu, 13 Mar 2014, Patrick Dowler wrote:

>
>
> On 13/03/14 03:31 PM, Petr Skoda wrote:
>> the biggest mistake of SIA 1.0 is there is not primary the concept of
>> filters built-in - I am just in troubles trying to access the multi
>> colour survey.
>> 
>> I do not believe the most use cases can be solved by stating
>> wavelengths in BAND....
>
>
> Are you talking about searching for data in the R band? You need to look at 
> the SIA-2.0 {query} capability; this defines query parameters for discovering 
> data. While it isn't perfect and it isn't trivial, one can write a numeric 
> query of the ObsCore model such that most of the returned data is in a band 
> very much like R. Applications could help users formulate such queries. If 
> you want to propose something more on the query side, please do so, but it 
> will probably not be considered in scope until SIA-2.1
>
> Or are you talking about creating a virtual R-band image out of a 3d cube? If 
> you are talking about a virtual image from a cube, the only kind of access 
> that is in scope for AccessData-1.0 is simple extraction is subsets. All 
> other kinds of operations are to be considered later in 1.x
>
> PS-It is by design that the numeric BAND parameter is the same in SIA and 
> AccessData, because they both specify a value or interval on the spectral 
> axis. However, the two capabilities do different things now and will diverge 
> further (in allowed parameters) in subsequent versions.
>
>
> -- 
>
> Patrick Dowler
> Canadian Astronomy Data Centre
> National Research Council Canada
> 5071 West Saanich Road
> Victoria, BC V9E 2E7
>
> 250-363-0044 (office) 250-363-0045 (fax)
>